of October 16, 2014.This information is current as

Cell MigrationCXCR4-Mediated Germinal Center CD4 T Follicular Dendritic Cell Regulation of

M. Druey and Gregory F. BurtonSariah A. Kell, Brandon F. Keele, Emily A. Palenske, Kirk Jacob D. Estes, Tyler C. Thacker, Denise L. Hampton,

http://www.jimmunol.org/content/173/10/6169doi: 10.4049/jimmunol.173.10.6169

2004; 173:6169-6178; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/173/10/6169.full#ref-list-1

, 32 of which you can access for free at: cites 65 articlesThis article

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2004 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

Follicular Dendritic Cell Regulation of CXCR4-MediatedGerminal Center CD4 T Cell Migration1

Jacob D. Estes,* Tyler C. Thacker,* Denise L. Hampton,‡ Sariah A. Kell,† Brandon F. Keele,*Emily A. Palenske,* Kirk M. Druey,‡ and Gregory F. Burton2†

Follicular dendritic cells (FDCs) up-regulate the chemokine receptor CXCR4 on CD4 T cells, and a major subpopulation ofgerminal center (GC) T cells (CD4�CD57�), which are adjacent to FDCs in vivo, expresses high levels of CXCR4. We thereforereasoned that GC T cells would actively migrate to stromal cell-derived factor-1 (CXCL12), the CXCR4 ligand, and tested thisusing Transwell migration assays with GC T cells and other CD4 T cells (CD57�) that expressed much lower levels of CXCR4.Unexpectedly, GC T cells were virtually nonresponsive to CXCL12, whereas CD57�CD4 T cells migrated efficiently despitereduced CXCR4 expression. In contrast, GC T cells efficiently migrated to B cell chemoattractant-1/CXCL13 and FDC super-natant, which contained CXCL13 produced by FDCs. Importantly, GC T cell nonresponsiveness to CXCL12 correlated with highex vivo expression of regulator of G protein signaling (RGS), RGS13 and RGS16, mRNA and expression of protein in vivo.Furthermore, FDCs up-regulated both RGS13 and RGS16 mRNA expression in non-GC T cells, resulting in their impairedmigration to CXCL12. Finally, GC T cells down-regulated RGS13 and RGS16 expression in the absence of FDCs and regainedmigratory competence to CXCL12. Although GC T cells express high levels of CXCR4, signaling through this receptor appearsto be specifically inhibited by FDC-mediated expression of RGS13 and RGS16. Thus, FDCs appear to directly affect GC T cellmigration within lymphoid follicles. The Journal of Immunology, 2004, 173: 6169–6178.

C ombating invading pathogens efficiently requires a seriesof orchestrated movements of T and B lymphocytes aswell as APCs to and within the secondary lymphoid tis-

sues (1). Chemokines are important proteins responsible for guid-ing lymphocytes into lymphoid organs and compartmentalizationwithin these sites (1, 2). Naive B and T cells enter the lymph nodethrough high endothelial venules (3–5), where the secondary lym-phoid tissue chemokine (CCL21) and the stromal cell-derived fac-tor-1 (CXCL12) chemokine are responsible for the induction offirm arrest of rolling T and B lymphocytes, respectively (6–9).After extravasation, naive lymphocytes are attracted to the para-cortical region of the parenchyma via CCL21 and EBI 1-ligandchemokine (CCL19) and CXCL12 (10).

Chemokines bind to receptors that belong to a large family ofseven-transmembrane, G protein-coupled receptors on the surfaceof leukocytes (11). Chemokine receptors are coupled to heterotri-meric G��� proteins. Agonist binding to the receptor catalyzes theexchange of GTP for GDP on the G� subunit and induces disso-ciation of the G� and G�� subunits (reviewed in Ref. 11). TheGTP-bound G� subunit and the G�� subunits independently ac-tivate multiple downstream effectors, one result of which is cell

migration (reviewed in Ref. 12). However, chemokine binding toits cognate receptor(s) does not always result in cell migration.Bleul et al. (13) showed that germinal center (GC)3 B cells, whichexpress high levels of surface CXCR4, failed to migrate in re-sponse to CXCL12 and that receptor desensitization may serve asa control mechanism of cell migration. Recently, the regulator ofG protein signaling (RGS) family of proteins was discovered andfound to associate with specific G� subunits, markedly stimulating(100- to 1000-fold) their native GTPase activity. As a result, thesignaling potential of the G protein-coupled receptor is inhibited(reviewed in Ref. 14). Because chemokine receptor signaling iscritical in directing cells to and within tissue compartments, local-ized control of RGS expression could contribute to receptor de-sensitization and prevent cells from migrating despite continuedexpression of receptors and exposure to chemokines (15).

A large body of work on the generation of thymus-dependent(TD) Ab responses suggests that Ag-signaled B cells initially un-dergo cognate interactions with T lymphocytes at the edge of thelymphoid follicles in secondary lymphoid tissue (16–20). Some Bcells differentiate at this stage into short-lived, Ab-forming cells,but other activated Ag-specific B and T cells migrate into the fol-licles where they form GCs surrounding follicular dendritic cells(FDCs) (16–18).

FDCs are restricted to the follicles of secondary lymphoid tissue(e.g., lymph nodes, spleen, and tonsils) and are thought to play akey role in the formation of GCs as well as in the selection, dif-ferentiation, and maintenance of memory B cells (21–23). Theirlong cytoplasmic extensions form a reticular network throughoutlymphoid follicles that traps and retains Ags in the form of im-mune complexes (24–27). FDC-trapped Ags persist for manymonths in an unprocessed form and serve to maintain long term,memory IgG and IgE responses to soluble protein Ags (21, 24, 25,

Departments of *Microbiology and Molecular Biology and †Chemistry and Biochem-istry, Brigham Young University, Provo, UT 84602; and ‡Molecular Signal Trans-duction Section, Laboratory of Allergic Diseases, National Institute of Allergy andInfectious Diseases, National Institutes of Health, Rockville, MD 20852

Received for publication July 8, 2004. Accepted for publication September 10, 2004.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by National Institutes of Health Grant AI39963 (toG.F.B.). J.D.E. was supported in part by a Brigham Young University Graduate Re-search Fellowship and the Vanice, Glen W., and Keith Reid Endowment for ScientificResearch at Brigham Young University.2 Address correspondence and reprint requests to Dr. Gregory F. Burton, Departmentof Chemistry and Biochemistry, Room C-211A BNSN, Brigham Young University,Provo, UT 84602. E-mail address: gburton@chem.byu.edu

3 Abbreviations used in this paper: GC, germinal center; ABS, Ab binding site; FDC,follicular dendritic cell; RGS, regulator of G protein signaling; TD, thymus dependent.

The Journal of Immunology

Copyright © 2004 by The American Association of Immunologists, Inc. 0022-1767/04/$02.00

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

28, 29). During humoral immune responses, FDCs play a criticalrole as accessory cells, presenting Ag-Ab complexes to B cells (27,30). FDCs also provide signals to T and B lymphocytes that altertheir state of activation/proliferation (31, 32) and render B cellsresponsive to chemoattractants (33). In addition, we have recentlyshown that FDCs up-regulate the chemokine receptor, CXCR4, onCD4 T cells in vitro and that CD4�CD57� GC T cells, a majorpopulation of GC T cells, which interact with FDCs in vivo (34),express high levels of CXCR4 compared with other CD4 T cells(CD57�) (35). Thus, in normal physiology, FDCs play an impor-tant role in the GC reaction that generates and maintains TD hu-moral immune responses.

Naive and resting B cells express the chemokine receptorCXCR5, which is required for the development of follicles in somesecondary lymphoid tissues as well as for B cell localization in thefollicles of spleen and Peyer’s patches (36). CXCR5 is also ex-pressed on a subset of circulating human memory CD4 T cells (37)and is up-regulated on some mouse CD4 T cells after immuniza-tion (38). On murine T cells, CXCR5 regulation appears to requireCD28-mediated, OX40 signaling (39). Recently, a novel subpopu-lation of B helper-T cells that are localized within lymphoid fol-licles has been defined and termed follicular B helper T cells orgerminal center Th cells (GC-Th) (40, 41). These CD4 T cellsexpress the CXCR5 chemokine receptor, are CD57�, and reside inthe GCs of secondary lymphoid tissues, where they produce ele-vated levels of IL-10 on stimulation and support the efficient pro-duction of IgG and IgA. Importantly, these GC T cells were foundto interact not only with GC B cells, but also with FDCs, as dem-onstrated by their direct contact with FDCs in vivo (34).

FDCs, in addition to their role in the development and mainte-nance of the GC reaction, may play a key role in recruiting both Band T cells into the lymphoid follicle to initiate the GC reaction. Insupport of this hypothesis, in situ hybridization analysis indicatesthat B cell chemoattractant-1 (CXCL13), a CXC chemokine thatattracts CXCR5� cells, is constitutively expressed by resident stro-mal cells in the GC (1, 42–44). These latter cells are most likelya subset of FDCs; however, it has not been shown that isolatedFDCs produce CXCL13. An understanding of chemokines andtheir interactions with specific receptors on motile B and follicularB helper T cells provides an explanation of how these cells comeinto contact with stationary FDCs to initiate the GC reaction (42).

In the present study, we examined the contributions of FDCs toGC T cell migration, and found that although GC T cells expresshigh levels of CXCR4, they were specifically nonresponsive toCXCL12-induced migration. Such nonresponsiveness correlatedwith FDC mediated up-regulation of RGS13 and RGS16 expres-sion in CD4 T cells. Additionally, we show that FDCs express andproduce CXCL13, which specifically attracts GC T cells viaCXCR5. FDC regulation of GC T cell responsiveness to CXCL12may play an important role in recruiting and retaining Ag-specificcells within the GC to induce and maintain the GC reaction, whichresults in the differentiation of Ag-specific B cells into Ab-formingcells and the generation of B memory cells (45).

Materials and MethodsFlow cytometric analysis

Cell surface Ags were detected using the following mAbs: anti-humanCD4-PC5 (13B8.2), anti-human CD14-PE (RM052), anti-human CD21-FITC (BL13), anti-human CD45RO-PE (UCHL1), anti-human CD57 con-jugated to FITC or biotin (NC1), and anti-CD69-PE (TP1.55.3; Immuno-tech, Westbrook, ME); anti-human CXCR4-PE (12G5; BD Biosciences,San Jose, CA); mouse IgM and anti-human FDC (HJ2; gift from Dr. M.Nahm, University of Alabama, Birmingham, AL); mouse IgG and anti-human CD21L (7D6; gift from Dr. Y.-J. Liu, DNAX, Palo Alto, CA); anddonkey, F(ab�)2, anti-mouse IgM-FITC (Jackson ImmunoResearch Labo-

ratories, West Grove, PA). Mouse isotype-matched IgG1 (679.1Mc7) andIgG2a (U7.27; Immunotech) were also used as controls. Cells were firstincubated for 30 min on ice with human ChromePure IgG (Jackson Im-munoResearch Laboratories) to block nonspecific Ab interactions and thenwith receptor-specific mAbs. Cells were then washed in cold PBS, and10,000 CD4 T cells were analyzed for immunofluorescence using anEPICS XL flow cytometer with EXPO32 ADC software (BeckmanCoulter, Fullerton, CA). Mouse isotype-matched control Abs were used todefine background fluorescence and establish positive and negative gating.Propidium iodide (0.5 �g; Sigma-Aldrich, St. Louis, MO) uptake was usedto exclude dead cells. Quantitation of Ab binding sites (ABS) was per-formed using Quantum Simply Cellular Microbeads (Sigma-Aldrich) ac-cording to the manufacturer’s instructions and as previously described (35).

FDC isolation

Human FDCs were isolated from tonsillar tissue as previously described(46). FDC-enriched preparations prepared using this procedure were ex-amined by flow cytometry and typically contained 75–90% FDCs withresidual cells consisting of T and B lymphocytes. In addition, in someexperiments FDCs were FACS-purified using a FACSVantage SEequipped with the FACSDiVa option and software (BD Biosciences, SanJose, CA). Briefly, low density tonsillar cells were collected from contin-uous Percoll gradients after centrifugation, washed, and incubated withheat-aggregated, human ChromePure IgG (Jackson ImmunoResearch Lab-oratories) to block nonspecific FcR binding, labeled with HJ2 (250 �l ofhybridoma supernatant) and 7D6 (250 �l of hybridoma supernatant) on icefor at least 1 h, followed by washing and addition of goat F(ab�)2 anti-mouse IgM-FITC (�-chain specific; 25 �g) and goat F(ab�)2 anti-mouseIgG-PE (�-chain specific, 25 �g). FDCs were sorted on HJ2hi/7D6hi events(this population typically ranged from 0.5–3% of the total population post-Percoll). In all coculture experiments, an FDC to CD4 T cell ratio of 1:10was used, because we found this to result in optimal FDC-lymphocyteinteractions (35). In some experiments FDCs were specifically depleted bycollecting the effluent from MACS columns used to positively select FDCs,and these cells were then subjected to an additional round of depletionusing HJ2 and magnetic beads (rat, anti-mouse IgM Dynabeads; DynalBiotech, Great Neck, NY) at a concentration of 10 beads per target cell.This treatment removed �90% of the FDCs.

CD4 T cell preparations

CD4�/CD57� GC T cells were isolated from human tonsillar tissue aspreviously described (35). Briefly, tonsils were cut into small sections, cellswere mechanically separated from tissue by repeat pipetting, and RBCswere removed by incubation for 5 min at room temperature in RBC-lysisbuffer (155 mM NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA). GC CD4 Tlymphocytes were purified by negative selection using a CD4� T cell iso-lation kit (Miltenyi Biotec, Auburn, CA), followed by positive selection byMACS using anti-CD57-biotin and streptavidin microbeads (Miltenyi Bio-tec). The resulting CD4�/CD57� preparations were �95% pure as as-sessed by flow cytometry.

Chemotaxis assay

Cell migration was evaluated using a 24-well, 5-�m pore size Transwellsystem (Costar, Cambridge, MA). Purified cells were washed once in che-motaxis medium (RPMI 1640 containing HEPES buffer (20 mM) and gen-tamicin (50 �g/ml; Invitrogen Life Technologies, Gaithersburg, MD) andthen adjusted to 5 � 106 cells/ml in the same medium. An aliquot (100 �l)of the above cell suspension containing 5 � 105 cells was placed on the topof the Transwell. Chemokines, prepared at the indicated concentrations(determined by titration assay) in chemotaxis medium (600-�l total vol-ume), were added to the bottom of the Transwell system. After 4- to 5-hincubation at 37°C in a 5% CO2 atmosphere, the inserts were removed, andthe number of cells that had migrated into the lower well was analyzed bycounting cells for 60 s on an EPICS XL flow cytometer (Beckman Coulter)with the gates set to acquire the particular cell of interest. In some exper-iments purified CD4� T cells were placed on top of the Transwell, and thecells that migrated through were labeled with anti-CD57-FITC and countedas described above. To establish the number of cells that migrated non-specifically, the migration assays were performed in parallel in the absenceof chemoattractant. Results are expressed as percent specific migration,which was calculated as follows: ((total number of cells migrating in thepresence of chemokine minus the total number of cells migrating in theabsence of chemokine) � the total number of input cells) � 100.

6170 FDC REGULATION OF GC T CELL MIGRATION

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

RNA isolation and RT-PCR analysis

Immediately upon isolation or in some cases after cell culture, an equalnumber of CD4 T cells was centrifuged and resuspended in PBS (100-�ltotal volume), and RNA STAT 60 (Tel-Test, Friendswood, TX) was addedat a ratio of 800 �l of RNA-STAT60 per 1.0 � 106 cells in sterile, 1.5-mlEppendorf tubes, after which the samples were stored at �80°C until test-ing. To each of the samples, 160 �l of chloroform was added per 800 �lof RNA-STAT 60 (Fisher Scientific, Pittsburgh, PA); the solution wasvortexed and then centrifuged at 13,000 � g for 15 min at 4°C. An equalvolume of the aqueous phase (400 �l) from each sample was transferred toa fresh tube containing the same volume of isopropanol, and the contentswere vortexed, incubated at �20°C for 30 min, and centrifuged at13,000 � g for 30 min at 4°C. After discarding the supernatant fluid, theRNA pellet was washed once in 75% ethanol (1.0 ml) and air-dried, and thesamples were resuspended to a volume of 20 �l in RNase/DNase-freewater (Invitrogen Life Technologies). Pretreatment of RNA with DNA-freeDNase (Ambion, Austin, TX) was performed according to the manufac-turer’s instructions to eliminate any potential DNA contamination. RNAfrom the same volume of sample from each tube (15 �l) was then reversetranscribed using the GeneAmp RNA PCR kit (Roche Molecular Systems,Branchburg, NJ) according to the manufacturer’s instructions. The result-ing cDNA was then PCR-amplified under the following conditions: onecycle at 94°C for 50 s, followed by 30 cycles of 94°C for 30 s, 58°C for1 min, 72°C for 1 min, and a single 10-min extension cycle at 72°C (47).In some instances (i.e., RGS1, -13, and -16), PCR analysis was performedfor 26, 28, and 30 cycles on aliquots of sample cDNA to ensure that anal-ysis occurred during the linear phase of the amplification process so thatsubtle differences in concentration would be detected. Densitometric anal-ysis of the resulting amplicons was performed using National Institutes ofHealth ImageJ software (Bethesda, MD). The following forward and re-verse primers were used: �-actin, 5�-CATCCTCACCCTGAAGTACC-3�and 5�-GGTGAGGATCTTCATGAGGT-3�, yielding a 398-bp amplicon(48); human CXCL13, 5�-TCATAGTCTGGAAGAAGAACAAGTCAA-3�and 5�-TCAGCATCAGGGAATCTTTCTCT-3� yielding a specific prod-uct of 143 bp with the primers spanning an intron (49); human RGS1,5�-CCCACATCTGGAATCTGGAA-3� and 5�-CTCTGCGCCTGGATAACTTT-3�, yielding a 620-bp amplicon; human RGS2, 5�-CCAAATCACCCCAAAAGCTGTCCTC-3� and 5�-CTCCTAGTCAGTTACTGGCTTCCTG-3�, yielding a 445-bp amplicon (47); human RGS13, 5�-ATGAGCAGGCGGAATTGTTGGA-3� and 5�-GAAACTGTTGTTGGACTGCATA-3�, yielding a 476-bp amplicon (50); and human RGS16,5�-TGGAGAGAGTCGTTCGACCTG-3� and 5�-TGTCCTCTTGCACTTGCTTTGC-3�, yielding a 535-bp amplicon (47). The PCR prod-ucts were separated on 1.5% agarose gels, stained in ethidium bromide,and photographed as negative images using the Fluor-S photographicMultiImager system with Quantity One image analysis software (Bio-Rad, Hercules, CA). To ensure specificity of human CXCL13, resolvedgels were transferred to Hybond-N� nylon transfer membrane (Amer-sham Biosciences, Arlington Heights, IL), and Southern blotting wasperformed using the human CXCL13 probe 5�-CCATTCAGCTTGAGGGTCCACACACA-3� (49) and the ECL detection system (Amer-sham Biosciences).

CXCL13 sandwich ELISA

The measurement of human CXCL13 from FDC and FDC-depleted cul-tured supernatants was performed using a specific sandwich DuoSetELISA kit (R&D Systems, Minneapolis, MN; detection limit, 25 pg/ml)according to the manufacturer’s instructions. Briefly, mouse, anti-humanCXCL13 mAb (2 �g/ml) was bound to each well of a 96-well microplate(Immulon 4 Plates; Dynatech Laboratories, Chantilly, VA) and used as thecapture Ab. A biotinylated, goat, anti-human CXCL13 polyclonal Ab (100ng/ml) was used as the detection Ab. Streptavidin-conjugated HRP wasused to detect bound biotin-labeled Ab and was visualized by incubationwith tetramethylbenzidine substrate for 20 min, followed by addition ofStop Solution (2 N H2SO4). Absorption was measured at 450 nm using aVmax Kinetic Microplate Reader (Molecular Devices, Sunnyvale, CA).The average of duplicate readings for each standard, control, and samplewas subtracted from the average reading obtained from noncoated wells. Astandard curve was generated with a four-parameter logistic curve fit usingSOFTmax PRO software (Molecular Devices).

Rabbit, anti-RGS16 (RGS16–210)

Affinity-purified, rabbit anti-human RGS16 was custom-produced for us byBio-Synthesis (Lewisville, TX). Rabbits were immunized with humanRGS16 peptide consisting of aa 38–53 (NH2-(GC)STGKFEWGSKHSKENR-COOH) conjugated to keyhole limpet hemocyanin. Serum was col-

lected and affinity-purified before use. The specificity of anti-RGS16–210was determined by Western blotting. For these experiments, His-tagged,recombinant RGS proteins were expressed in bacteria and purified as pre-viously described (51–53).

Immunohistochemistry

Tonsils obtained from elective surgery were fixed in Streck Tissue Fixative(Streck Laboratories, La Vista, NE) at room temperature for 1–3 days, afterwhich they were embedded in paraffin and cut into 6-�m sections forimmunohistochemical staining. The sections were dewaxed and rehydratedthrough graded ethanol treatments, then washed in double-distilled waterfor 5 min. Before ethanol treatments, sections were subsequently exposedto 0.6% hydrogen-peroxide-ethanol solution to quench endogenous perox-idase activity. After a 2-h preincubation with 5% normal goat serum (Vec-tastain ABC kit; Vector Laboratories, Burlingame, CA) in TNB (0.10 MTris-HCl, 0.15 M NaCl, and 0.5% blocking reagent (NEN, Boston, MA)),sections were layered overnight with primary Abs colocalized on the samesections for double-label immunohistochemistry. Mouse anti-human CD57(IgM; 1 �g/ml; NeoMarkers, Fremont, CA) and rabbit, affinity-purifiedRGS13 (provided by Dr. J. H. Kehrl, National Institute of Allergy andInfectious Diseases, National Institutes of Health, Bethesda, MD), rabbit,affinity-purified RGS16-210, or preimmune rabbit IgG polyclonal Abswere used at 1 �g/ml in TNB with 5% normal goat serum. Detection of therabbit primary Abs (anti-human RGS13 and RGS16, and preimmune rabbitIgG) was performed after incubation with biotinylated goat anti-rabbit IgGAb (4 �g/ml; Vector Laboratories) for 30 min at room temperature, fol-lowed by incubation with alkaline phosphatase-labeled avidin-biotin re-agent (ABC Elite kit; Vector Laboratories) for 30 min at room temperatureand visualized using Vector Red as the substrate for 30 min in a humidchamber at room temperature in the dark. Detection of the mouse anti-human CD57 primary Ab was performed after incubation with biotinylatedgoat anti-mouse IgM (�-chain specific) Ab (1 �g/ml; Vector Laboratories)for 30 min at room temperature, followed by incubation with HRP-labeledavidin-biotin reagent (ABC Elite kit) for 30 min at room temperature andvisualized with Vector SG (bluish-gray color; Vector Laboratories) for 10min in a humid chamber at room temperature in the dark. To avoid cross-reaction between the detection system for the first and second mAbs, bi-otinylated goat anti-mouse IgM Ab was used at a lower concentration inthe second step of the double-staining protocol, and controls were alsoperformed omitting the second primary Ab (CD57) or replacing it with anisotype-matched control mAb with irrelevant specificity.

Statistical analysis

All data are presented as the mean � SEM of duplicate or triplicate sam-ples and represent at least three independent experiments. Analysis wasperformed using Student’s t test. A value of p � 0.05 was consideredsignificant.

ResultsCXCR4hi GC T cells migrate poorly to CXCL12

We previously reported that FDCs up-regulate CXCR4 on CD4 Tcells and that a major population of GC CD4 T cells that bearCD57 and interact with FDCs in vivo also expresses high levels ofCXCR4 (35). Because human T and B cells can migrate to FDCs(54), we hypothesized that CXCR4 may be involved in this processand sought to test this postulate. Because GC T cells bear up to6-fold more CXCR4 than other CD4 T cells, we examined themigratory capacity of these cells to the CXCR4 ligand, CXCL12,and compared this to the migration of other CD4 T cells (CD57�)from the same tissue. Remarkably, although GC T cells expressed�4-fold more CXCR4 than CD57�CD4 T cells from the sametissue (Fig. 1A), they migrated much less efficiently to CXCL12than did the CD57�CD4 T cells (Fig. 1B). Because our GC T cellisolation procedure used positive selection for CD57, we excludedthe possibility that anti-CD57 binding altered the migratory capac-ity of the cells to CXCL12. We negatively selected CD4 T cellsand then phenotyped the cells after migration through the Trans-well membrane to chemoattractant. These GC T cells were alsononresponsive to CXCL12 compared with CD57�CD4 T cells(data not shown).

6171The Journal of Immunology

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

FDCs express CXCL13, a potent chemoattractant for GC T cells

We next asked whether the nonresponsiveness of the GC T cellswas specific to CXCL12 or whether the cells exhibited a general-ized loss of migratory ability. Because GC T cells express CXCR5(40), we analyzed their migration to the CXCR5 ligand, CXCL13.GC T cells expressed �10-fold higher levels of CXCR5 than otherCD4 T cells (Fig. 2A) and migrated 10-fold more efficiently to thechemokine CXCL13 compared with CXCL12 (Fig. 2B). In con-trast, although CD57�CD4 T cells once again migrated efficientlyto CXCL12, they failed to migrate to CXCL13 (Fig. 2B). Theseresults highlight a differential responsiveness of GC and non-GC Tcells to CXCL12 and CXCL13.

Because FDCs produce chemoattractants that specifically attractCD4 T cells (54), we next evaluated migration of purified CD4 Tcell preparations from tonsils to supernatant fluid obtained fromcultured FDCs. Interestingly, GC T cells, but not CD57�CD4 Tcells, migrated efficiently to FDC supernatant (Fig. 3A), but neithercell migrated to supernatant fluid from control cultures specificallydepleted of FDCs. Because FDC supernatant served as a chemoat-tractant for GC T cells, and in situ hybridization suggests thatCXCL13 is constitutively expressed by FDCs (1, 42–44), we de-termined whether GC T cell migration to FDCs was mediated byCXCL13 and CXCR5. A blocking CXCL13 Ab was added to su-pernatant fluid from cultured FDCs, and CD4 T cell migration was

assessed as before (Fig. 3B). Although the addition of anti-CXCL13 to FDC supernatant failed to completely block GC T cellmigration, it reduced it by �50%, whereas incubation of GC Tcells with anti-CXCR5 before the migration assay completely ab-lated migration to the same FDC supernatant (Fig. 3C). These dataindicate that migration of GC T cells toward FDCs was mediatedby CXCR5 in response to FDC-derived CXCL13. To test whetherFDCs expressed CXCL13, we performed RT-PCR analysis ofMACS-enriched and FACS-purified FDCs (Fig. 4A). FDC super-natant was assessed for CXCL13 by ELISA (Fig. 4B). FDCs dem-onstrated robust expression of CXCL13 mRNA, whereas tonsillarcells specifically depleted of FDCs showed little expression. Inaddition, FDCs produced between 61 and 866 pg/ml CXCL13,depending on the FDC donor, whereas tonsillar cells depleted ofFDCs did not secrete this chemokine. These data indicate thatFDCs produce CXCL13 and are probably the cell identified in insitu hybridization studies (1, 42–44).

Differential RGS gene expression by GC T cells correlates toimpaired migration to CXCL12

The fact that GC T cells migrated efficiently toward CXCL13, butfailed to migrate to CXCL12, suggested that these cells might bedesensitized to signaling through CXCR4 even though their sur-face expression of this receptor was much higher than that of otherCD4 T cells from the same tissue. To analyze the mechanism of

FIGURE 1. GC T cells are nonresponsive to CXCL12-induced migra-tion, although they express high levels of CXCR4. A, CXCR4 was mea-sured on isolated CD57�CD4 T cells and GC T cells, and CXCR4 receptornumber (ABS) was determined. GC T cells express �4-fold higher surfaceexpression of CXCR4 compared with the CD57� T cells. B, PurifiedCD57�CD4 T cell and GC T cell migration to CXCL12 was determinedafter 5 h of culture in a Transwell migration system, after which the numberof live cells in the lower chamber was counted. The data represent themean � SEM of triplicate wells. �, p � 0.01. The data are representativeof six independent experiments.

FIGURE 2. GC T cells migrate efficiently to CXCL13. A, CXCR5 re-ceptor number (ABS) was determined on GC and CD57�CD4� T cells.GC T cells expressed higher levels of CXCR5 compared with CD57� Tcells. B, GC T cells and CD57�CD4 T cells were purified, and their mi-gratory capacities to CXCL12 (1000 ng/ml) and CXCL13 (1000 ng/ml)were determined as before. Note that the GC T cells were fully capable ofmigrating to CXCL13. The data represent the mean � SEM of triplicatewells. �, p � 0.01. The data are representative of six independentexperiments.

6172 FDC REGULATION OF GC T CELL MIGRATION

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

GC T cell nonresponsiveness to CXCL12, we compared the ex-pression of several RGS genes in GC T cells and CD57�CD4 Tcells. Because GC B cells highly express RGS1 and RGS13 and arealso nonresponsive to CXCL12-mediated migration (50, 55), wereasoned that these genes may also be differentially expressed in

GC T cells compared with CD57�CD4 T cells. We also analyzedthe expression of two other genes expressed in T cells, RGS16 andRGS2. Interestingly, GC T cells expressed high levels of RGS13and RGS16 mRNA compared with CD57�CD4 T cells from thesame tissue, whereas there was no difference in RGS1 mRNA ex-pression (Fig. 5A). Furthermore, no difference was detected inRGS2 mRNA in the T cell populations. To ensure that our PCRamplification was performed during the linear phase of the ampli-fication process, when subtle differences in RGS expression couldbe more readily detected, we examined amplification productsfrom cycles 26, 28, and 30 and subjected them to densitometry(Fig. 5B). This approach confirmed the presence of higher levels ofRGS13 and RGS16 mRNA expression in GC T cells, whereasRGS1 expression was equivalent among the cells at a given cycle.In a preliminary experiment we confirmed previous findings thatGC B cells expressed high levels of RGS13 mRNA (50) and ob-served that these lymphocytes also expressed RGS16 mRNA, al-beit at 6-fold lower levels than RGS13, whereas other B cells didnot express detectable levels of either of these transcripts after 30cycles of amplification (data not shown). To establish RGS13 andRGS16 protein expression in GC T cells, we obtained a previously

FIGURE 3. FDCs attract GC T cells via production of CXCL13. A,FDCs and FDC-depleted cells were purified and cultured in complete me-dium supplemented with 10% FBS for 6 days. Tissue culture supernatantfrom these cells was harvested, filtered through a 0.22-�m pore size filter,and used as the only source of chemoattractant in Transwell migrationstudies with purified GC T cells or CD57� T cells. FDC supernatant at-tracted GC T cells, but not other CD4 T cells, whereas FDC-depletedsupernatant did not attract any tonsillar CD4 T cells. B, FDC supernatantwas pretreated with anti-CXCL13 (20 �g/ml) for 30 min at 37°C and usedin Transwell migration assays with purified GC T cells. The addition ofanti-CXCL13 reduced GC T cell migration by �2-fold. C, GC T cells werepretreated with either anti-CXCR5 (10 �g) or control Ab (10 �g) for 30min on ice before performing the migration studies with FDC supernatant.Addition of anti-CXCR5 abrogated migration to the FDC supernatant. Thedata represent the mean � SEM of triplicate wells. �, p � 0.03. The dataare representative of at least three independent experiments.

FIGURE 4. FDCs express CXCL13. A, mRNA was isolated from equalnumbers of either MACS-enriched FDCs and FDC-depleted cells orFACS-purified FDCs and FDC-negative cells (cells not labeling with FDC-specific mAbs), followed by RT-PCR for �-actin and CXCL13. Southernblotting was performed using a labeled probe specific for CXCL13 anddeveloped by ECL. Note the strong hybridization signal from FDCs, butnot from FDC-depleted or negative cells. The data are representative ofthree independent experiments. B, CXCL13-specific sandwich ELISA wasperformed on five independent samples of supernatant fluid obtained fromFDC cultures to determine the amount of CXCL13 produced. Culture ofFDC-depleted cells did not detect the presence of CXCL13. The data rep-resent the mean � SEM of two or more replicates of 6-day cultured su-pernatant. �, p � 0.01.

6173The Journal of Immunology

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

characterized RGS13-specific Ab (provided by Dr. J. Kehrl, NationalInstitute of Allergy and Infectious Diseases) (50) and generated a newanti-RGS16 Ab (RGS16–210). Western blotting experiments con-firmed the specificity of RGS16–210 (Fig. 6A). Immunohisto-chemistry was then performed on tonsil sections to determine theexpressions of RGS13 and RGS16 (Fig. 6B) in vivo. RGS13 ex-pression was largely confined to lymphoid follicles, and both GCT cells (CD57�) and other cells were labeled. Some interfollicularlabeling was also observed. RGS16 expression was present in GCT cells and other follicular cells, but, in contrast to RGS13, wasalso found expressed in tonsillar epithelium.

Because GC T and B cells exist adjacent to FDCs in secondarylymphoid tissue and exhibit similar patterns of RGS13 and RGS16gene expression, we hypothesized that FDCs could induce the ex-pression of these genes in CD57�CD4 T cells, rendering themunresponsive to CXCL12. When CD57�CD4 T cells (which didnot express RGS13 or RGS16) were cultured with FDCs for 24 h,both RGS13 and RGS16 mRNA were up-regulated (Fig. 7A); how-ever, no changes were observed in their surface expression ofCXCR5 or their ability to migrate to CXCL13 (data not shown).The expression of RGS13 and RGS16 was not up-regulated in short

term cultures if the FDC and T cells were separated by a semi-permeable Transwell membrane (Fig. 7A). Coculture of FDCs withCD57�CD4 T cells significantly impaired migration to CXCL12(Fig. 7B). Thus, FDCs modulate RGS13 and RGS16 gene expres-sion directly, regulating CD4 T cell migration to CXCL12. There-fore, GC T cells cultured in the absence of FDCs should down-regulate RGS13 and RGS16 to restore responsiveness to CXCL12.GC T cells cultured in the absence of FDCs for 24 h down-regu-lated both RGS13 and RGS16 mRNA (Fig. 8A), and this expres-sion correlated directly with enhanced migration to CXCL12 (Fig.8B). Furthermore, the specific migration of these GC T cells wasequivalent to that of freshly isolated CD57�CD4 T cells. Thesedata are consistent with the hypothesis that FDC-CD4 T cell in-teractions occur in vivo and that these interactions result inchanges in RGS13 and RGS16 gene expression that result in GC Tcell nonresponsiveness to CXCL12.

DiscussionLong term TD Ab responses, with high affinity IgG production, arecritically dependent on a productive GC reaction, involving acti-vated, Ag-specific B and T cells surrounding Ag-bearing FDCs(20, 27, 29, 56). Cellular localization of Ag-specific lymphocyteswithin lymphoid follicles is controlled by chemokines and the ex-pression of chemokine receptors on these cells (1, 2). BecauseFDCs up-regulate CXCR4 on CD4 T cells, and GC T cells expresshigh levels of CXCR4 (35), these cells should migrate efficiently toCXCL12. However, we observed that although GC T cells ex-pressed high levels of CXCR4, they were nonresponsive toCXCL12. Such nonresponsiveness could be explained by in-creased expression of RGS13 and RGS16 in this cell population.When GC T cells were incubated in the absence of FDCs, theyrapidly down-regulated RGS13 and RGS16 expression and becameresponsive to CXCL12. Additionally, incubation of CD57�CD4�

T cells with FDCs resulted in increased expression of RGS13 andRGS16, and this expression correlated with impaired CXCR4-evoked migration. Thus, the differential expression of RGS13 andRGS16 in GC T cells probably accounts for the observed CXCR4desensitization. Our data also suggest that FDC contact with GC Tcells in vivo results in inhibited migration to CXCL12 and thatspatial separation from FDCs is required for restoration ofCXCL12 responsiveness.

The nature of the FDC signal(s) responsible for up-regulationof RGS13 and RGS16 and impaired migration to CXCL12 iscurrently under investigation. Although the specific moleculesinvolved in this signaling pathway have not yet been identified,our data suggest that FDC contact may be required to blockmigration to CXCL12, as was evidenced by the ablation of thiseffect when FDCs were separated from CD4 T cells by a Trans-well in short term cultures. Although it is possible that our shortterm cultures did not allow the production of sufficient solublesignal to mediate RGS expression, Yuda et al. (34) found thatCD57� GC T cells directly contact FDCs in secondary lym-phoid tissue in vivo. The failure of GC T cells to migrate toCXCL12 is reminiscent of the inability of GC B cells to migrateto this chemokine despite high levels of CXCR4 expression(13). Interestingly, when GC B cells further differentiate, theyregain responsiveness to CXCL12. This chemokine is expressedin areas surrounding GCs and in the bone marrow, and thedifferentiated B cells leave the GC by migrating towardCXCL12 (reviewed in Ref. 57). GC B cell nonresponsiveness toCXCL12 may be controlled by RGS1 and RGS13 expression(47, 50). In GC T cells, however, it is unlikely that RGS1 has asignificant function in regulating CXCL12-induced migration,

FIGURE 5. GC T cells express high levels of RGS13 and RGS16mRNA compared with CD57� T cells. A, Equal numbers of GC T cells andCD57� T cells were purified, their mRNA were isolated, and 30 cycles ofRT-PCR were performed, followed by electrophoresis and visualization. B,cDNA samples were reamplified for 26, 28, or 30 cycles and examined asdescribed above to ensure that comparisons of GC T cells and other CD4�

T cells were made during the linear portion of the amplification process.Note the augmentation of RGS13 (35-, 71-, and 4-fold higher amplicondensity at cycles 26, 28, and 30, respectively) and RGS16 mRNA expres-sion (4-, 5-, and 2-fold higher density) in GC T cells compared withCD57� T cells, whereas RGS1 expression was equivalent (amplicon den-sity of 1 for cycles 26, 28, and 30). The data represent mRNA expressionfrom two donors and are representative of six independent experiments.

6174 FDC REGULATION OF GC T CELL MIGRATION

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

because RGS1 mRNA was equivalent in migrating and nonmi-grating CD4 T cell populations (Fig. 5). In contrast,CD57�CD4 T cells that migrated efficiently to CXCL12 ex-pressed very low levels (often undetectable) of RGS13 andRGS16 mRNA, whereas GC T cells that were refractory toCXCL12 migration expressed high levels of these RGS genes(Fig. 5). Shi et al. (50) demonstrated RGS13 expression in cellslocated in the light zone of the GC, which is the site occupiedby both GC B and T cells adjacent to FDCs. In addition, RGS13

strongly impairs signaling through G�i-regulated pathways, in-cluding those evoked by CXCR4 and CXCR5. Because GC Tcells migrated efficiently to the CXCR5 ligand, CXCL13,RGS13 may preferentially inhibit CXCR4-mediated signaling.Although selectivity of RGS proteins for certain receptors hasnot been well established, some data suggest that this phenom-enon may occur. RGS2, but not RGS4, binds directly to thethird intracellular loop of specific muscarinic receptor subtypes(58). T cells from mice that transgenically overexpress RGS16

FIGURE 6. CD57� GC T cells express RGS13and RGS16 proteins in vivo. A, Specificity of anti-RGS16 Abs. Recombinant, purified RGS proteins(200 ng) were separated by SDS-PAGE and immu-noblotted with preimmune antiserum (2 �g/ml) oranti-RGS16–210 (1 �g/ml) in buffer alone or pre-incubated with either recombinant RGS16 (rRGS16)or GST (200 nM). Coomassie blue staining was usedto verify the integrity of the electrophoresed proteins(bottom panel; 5 �g of each protein). B, Immuno-histochemical localization of RGS13- and RGS16-expressing cells in tonsils. Tonsillar tissue sectionswere stained with mouse anti-CD57 (dark blue) andpreimmune, rabbit IgG (a and b), rabbit anti-RGS13(c and d), or rabbit anti-RGS16–210 (e and f; allred). Preimmune rabbit IgG yielded no staining (aand b); however, high magnification images showthat most CD57� GC T cells express both RGS13(d) and RGS16 (f; black arrows). Other cells withinGCs also expressed both RGS13 and RGS16, whichprobably represent B cell or potentially FDC expres-sion. Whereas RGS13 expression was mostly con-tained within GCs, with some expression in cells inthe interfollicular region (c), considerable expres-sion of RGS16 was localized within intraepitheliallymphocytes lining the tonsillar crypts (e). The datarepresent images obtained from three different tissuedonors. TE, tonsillar epithelia; TZ, T cell zone.

6175The Journal of Immunology

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

exhibit impaired migration to CXCL12, but not CCL21 orCCL7 (59).

B cell differentiation signals may be important cues for egress ofthese cells from the GC to the bone marrow, where their differ-entiation into plasma cells is completed, and the bulk of Ab pro-duction occurs. Because it has been reported that B cell migrationto CXCL12 is significantly reduced initially after BCR engage-ment (13), we hypothesize that GC B cell contact with Ag-bearingFDCs leads to the induction of CXCL12 nonresponsiveness. Asubsequent signaling event must occur during the GC reaction thatrestores B cell migratory competence and allows them to leave theGC. Perhaps the release of immune complex-coated bodies (icco-somes) from FDCs and their presentation to GC T cells by GC Bcells (which can occur away from FDCs) result in the differenti-ation signals needed for plasma cell development (45, 60, 61) andthe signals that restore migratory competence to CXCL12. Regard-less of whether GC B cell presentation of Ag to GC T cells also

removes the block to GC T cell migration is unknown, but it iscurrently an area of active investigation. Our data indicate thatremoval of GC T cells from FDCs is required for restoration ofmigratory competence to CXCL12 because GC T cells cultured inthe absence of FDCs resulted in decreased expression of RGS13and RGS16 and migration to CXCL12 (Fig. 7). It will be interest-ing to determine whether GC T cells can overcome their nonre-sponsiveness to CXCL12 similar to differentiating GC B cells and,if so, to determine to which tissue sites these T cells traffic (i.e.,nonlymphoid or secondary lymphoid tissues).

GC T cells, but not CD57�CD4 T cells, migrated to FDC su-pernatant (Fig. 3), and Ab blocking studies indicated that FDC-secreted CXCL13 mediated this movement. Although the reasonwhy Ab to CXCR5 was more effective than anti-CXCL13 inblocking migration is not known, it may relate to differences inaffinity between the Ab preparations. Previous work indicates thatFDCs produce chemoattractant factors resulting in T cell cluster-ing around FDCs (54). Our data indicate that FDCs attracted CD4T cells, specifically GC T cells, through production of CXCL13,and that purified FDCs, but not other cells from the same tissue,were the major producers of this chemokine. These findings arealso consistent with earlier in situ hybridization and immunohis-tochemical studies, indicating a reticular pattern of expression ofCXCL13 throughout primary lymphoid follicles in murine spleens,lymph nodes, and Peyer’s patches. This expression pattern was

FIGURE 7. FDCs up-regulate the expression of RGS13 and RGS16 inCD57�CD4 T cells, which correlates with impaired CXCL12 migration. A,Purified CD57�CD4 T cells were cultured alone, with FDCs, or separatedfrom FDCs by a 0.4-�m pore size Transwell filter. The FDC control rep-resents RNA isolated from CD57� T cells cultured alone for 24 h pooledwith FDCs cultured alone for this same period of time. This control indi-cates that FDCs were not the source of the expressed RGS mRNA in thecoculture experiment. The data represent mRNA expression from one do-nor and are representative of three independent experiments. B, PurifiedCD57�CD4 T cells were cultured alone or with FDCs for 24 h, and then5 � 105 cells were placed in a Transwell migration assay as describedabove using CXCL12 as chemoattractant. The FDC control consists ofFDCs and CD4�CD57� T cells cultured separately and then combinedimmediately before the migration assay. This control ensures that T cellmigration to CXCL12 is not impaired by FDC production of CXCL13 orphysical inhibition of T cell migration. The data represent the mean �SEM of triplicate wells. �, p � 0.0001. The data are representative of atleast two independent experiments. Note that CD57�CD4 T cell migrationwas reduced by 8-fold when cocultured with FDCs.

FIGURE 8. GC T cells down-regulated RGS13 and RGS16 mRNA andregained responsiveness to CXCL12 when cultured in the absence ofFDCs. A, RNA was isolated from freshly isolated, purified GC T cells orfrom GC T cells cultured in the absence of FDCs for 24 h. The datarepresent mRNA from one donor and are representative of two independentexperiments. B, Freshly isolated GC T cells were incubated either at 4 orat 37°C in the absence of FDCs for 24 h, then assessed for their migratoryability to CXCL12 as described above. The data represent the mean �SEM of triplicate wells. �, p � 0.01. The data are representative of at leasttwo independent experiments.

6176 FDC REGULATION OF GC T CELL MIGRATION

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

also apparent in human secondary lymphoid tissues, suggestingthat FDCs (which are organized in a reticulum) are the major pro-ducers of CXCL13 (43, 49, 62). Thus, FDCs appear to not onlyprovide Ag for maintenance of long term Ab responses, but also tosecrete a chemokine, CXCL13, that attracts Ag-specific B and Tcells into the lymphoid follicle to initiate the GC reaction andretain them in this location (through CXCL12 nonresponsiveness).

It seems paradoxical that FDCs up-regulate CXCR4 expressionin GC T lymphocytes (35), yet induce RGS expression to inhibitCXCR4-evoked signaling. We reason that regulation of CXCL12responsiveness during the first phase of the GC reaction may serveto keep Ag-specific B and T cells within GCs to induce Ab-form-ing cell generation (45, 57). We envision that FDCs secreteCXCL13 to attract Ag-specific B and T cells from the site of initialactivation by dendritic cells in the TD zones of secondary lym-phoid tissues. Although the frequency of Ag-specific B and T cellsis low, GC development requires both cell types to be present.Premature migration would probably result in the inability of theGC reaction to progress, resulting in an inefficient response to TDAgs. Therefore, up-regulation of CXCR4, followed by down-reg-ulation of its signaling capacity (until additional signals are pro-vided), would allow Ag-specific cells to be retained in the GC forsufficient time periods to induce the generation of Ab-formingcells. Thereafter, when the majority of memory B cells are gener-ated, the nature and type of T cell signals would be altered.

In addition to contributing to the GC reaction, another poten-tially important consequence of FDC inhibition of GC T cell mi-gration out of the follicle relates to HIV pathogenesis. FDCs are along term repository of infectious virus trapped early in HIV in-fection (46, 63–65). FDC inhibition of GC T cell migration may beimportant in potentiating HIV transmission in these sites. In sup-port of this hypothesis, we have shown previously that CXCR4hi-expressing GC T cells are highly susceptible to infection by X4isolates of HIV (35). Hufert et al. (66) demonstrated that in HIV-infected subjects, these cells have up to a 10-fold higher frequencyof infection than other cells, and that active viral replication wasdetected almost exclusively in CD4�CD57� GC T cells. There-fore, FDC-mediated inhibition of GC T cell migration from thefollicle would keep highly susceptible cells in a microenvironmentextremely conducive to HIV infection and replication, therebycontributing to the disease state. A further understanding of FDCcontributions to the GC microenvironment may be important inboth understanding and eventually regulating humoral immunityas well as HIV pathogenesis.

AcknowledgmentsWe acknowledge the assistance of Dr. Kipp M. Robins, Dr. Randal B. Gibb,and their associates at the Utah Valley Regional Medical Center, HealthSouth Provo Surgical Center, and Central Utah Surgical Center for pro-viding tissues. We also acknowledge reagents supplied by the NationalInstitutes of Health, AIDS Research and Reference Reagent Program(Rockville, MD).

References1. Cyster, J. G. 1999. Chemokines and cell migration in secondary lymphoid organs.

Science 286:2098.2. Baggiolini, M. 1998. Chemokines and leukocyte traffic. Nature 392:565.3. Mackay, C. R., W. L. Marston, and L. Dudler. 1990. Naive and memory T cells

show distinct pathways of lymphocyte recirculation. J. Exp. Med. 171:801.4. Butcher, E. C., and L. J. Picker. 1996. Lymphocyte homing and homeostasis.

Science 272:60.5. Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte

emigration: the multistep paradigm. Cell 76:301.6. Gunn, M. D., K. Tangemann, C. Tam, J. G. Cyster, S. D. Rosen, and

L. T. Williams. 1998. A chemokine expressed in lymphoid high endothelialvenules promotes the adhesion and chemotaxis of naive T lymphocytes. Proc.Natl. Acad. Sci. USA 95:258.

7. Campbell, J. J., J. Hedrick, A. Zlotnik, M. A. Siani, D. A. Thompson, andE. C. Butcher. 1998. Chemokines and the arrest of lymphocytes rolling underflow conditions. Science 279:381.

8. Pachynski, R. K., S. W. Wu, M. D. Gunn, and D. J. Erle. 1998. Secondarylymphoid-tissue chemokine (SLC) stimulates integrin �4�7-mediated adhesion oflymphocytes to mucosal addressin cell adhesion molecule-1 (MAdCAM-1) underflow. J. Immunol. 161:952.

9. Stein, J. V., A. Rot, Y. Luo, M. Narasimhaswamy, H. Nakano, M. D. Gunn,A. Matsuzawa, E. J. Quackenbush, M. E. Dorf, and U. H. von Andrian. 2000. TheCC chemokine thymus-derived chemotactic agent 4 (TCA-4, secondary lym-phoid tissue chemokine, 6Ckine, exodus-2) triggers lymphocyte function-associ-ated antigen 1-mediated arrest of rolling T lymphocytes in peripheral lymph nodehigh endothelial venules. J. Exp. Med. 191:61.

10. Cyster, J. G. 1999. Chemokines and the homing of dendritic cells to the T cellareas of lymphoid organs. J. Exp. Med. 189:447.

11. Jung, S., and D. R. Littman. 1999. Chemokine receptors in lymphoid organ ho-meostasis. Curr. Opin. Immunol. 11:319.

12. Ward, S. G., K. Bacon, and J. Westwick. 1998. Chemokines and T lymphocytes:more than an attraction. Immunity 9:1.

13. Bleul, C. C., J. L. Schultze, and T. A. Springer. 1998. B lymphocyte chemotaxisregulated in association with microanatomic localization, differentiation state,and B cell receptor engagement. J. Exp. Med. 187:753.

14. De Vries, L., B. Zheng, T. Fischer, E. Elenko, and M. G. Farquhar. 2000. Theregulator of G protein signaling family. Annu. Rev. Pharmacol. Toxicol. 40:235.

15. Kehrl, J. H. 1998. Heterotrimeric G protein signaling: roles in immune functionand fine-tuning by RGS proteins. Immunity 8:1.

16. MacLennan, I. C. 1994. Germinal centers. Annu. Rev. Immunol. 12:11.17. Tarlinton, D. 1998. Germinal centers: form and function. Curr. Opin. Immunol.

10:245.18. Liu, Y. J., and J. Banchereau. 1997. Regulation of B-cell commitment to plasma

cells or to memory B cells. Semin. Immunol. 9:235.19. Rajewsky, K. 1996. Clonal selection and learning in the antibody system. Nature

381:751.20. Kelsoe, G. 1996. Life and death in germinal centers (redux). Immunity 4:107.21. Tew, J. G., R. P. Phipps, and T. E. Mandel. 1980. The maintenance and regulation

of the humoral immune response: persisting antigen and the role of follicularantigen-binding dendritic cells as accessory cells. Immunol. Rev. 53:175.

22. Hannum, L. G., A. M. Haberman, S. M. Anderson, and M. J. Shlomchik. 2000.Germinal center initiation, variable gene region hypermutation, and mutant B cellselection without detectable immune complexes on follicular dendritic cells.J. Exp. Med. 192:931.

23. Klaus, G. G. B., J. H. Humphrey, A. Kunkl, and D. W. Dongworth. 1980. Thefollicular dendritic cell: its role in antigen presentation in the generation of im-munological memory. Immunol. Rev. 53:3.

24. Mandel, T. E., R. P. Phipps, A. Abbot, and J. G. Tew. 1980. The folliculardendritic cell: long term antigen retention during immunity. Immunol. Rev. 53:29.

25. Szakal, A. K., M. H. Kosco, and J. G. Tew. 1989. Microanatomy of lymphoidtissue during the induction and maintenance of humoral immune responses: struc-ture function relationships. Annu. Rev. Immunol. 7:91.

26. Szakal, A. K., K. L. Holmes, and J. G. Tew. 1983. Transport of immune com-plexes from the subcapsular sinus to lymph node follicles on the surface ofnonphagocytic cells, including cells with dendritic morphology. J. Immunol.131:1714.

27. Tew, J. G., M. H. Kosco, and A. K. Szakal. 1989. The alternative antigen path-way. Immunol. Today 10:229.

28. Helm, S. L., G. F. Burton, A. K. Szakal, and J. G. Tew. 1995. Follicular dendriticcells and the maintenance of IgE responses. Eur. J. Immunol. 25:2362.

29. Klaus, G. G. B., and A. Kunkl. 1982. The role of T cells and B cell priming andgerminal centre development. Adv. Exp. Med. Biol. 149:743.

30. Tew, J. G., M. H. Kosco, G. F. Burton, R. M. DiLosa, and A. K. Szakal. 1990.Follicular dendritic cells and antigen presentation. In Dendritic Cells in LymphoidTissues. Y. Imai, J. G. Tew, and E. C. M. Hoefsmit, eds. Elsevier, Amsterdam,p. 111.

31. Kosco, M. H., E. Pflugfelder, and D. Gray. 1992. Follicular dendritic cell-depen-dent adhesion and proliferation of B cells in vitro. J. Immunol. 148:2331.

32. Burton, G. F., D. H. Conrad, A. K. Szakal, and J. G. Tew. 1993. Folliculardendritic cells (FDC) and B cell co-stimulation. J. Immunol. 150:31.

33. Burton, G. F., L. I. Kupp, E. C. McNalley, and J. G. Tew. 1995. Folliculardendritic cells and B cell chemotaxis. Eur. J. Immunol. 25:1105.

34. Yuda, F., K. Terashima, M. Dobashi, M. Ishikawa, and Y. Imai. 1989. Ultra-structural analysis of HNK-1� cells in human peripheral blood and lymph nodes.Histol. Histopathol. 4:137.

35. Estes, J. D., B. F. Keele, K. Tenner-Racz, P. Racz, M. A. Redd, T. C. Thacker,Y. Jiang, M. J. Lloyd, S. Gartner, and G. F. Burton. 2002. Follicular dendriticcell-mediated up-regulation of CXCR4 expression on CD4 T cells and HIVpathogenesis. J. Immunol. 169:2313.

36. Forster, R., A. E. Mattis, E. Kremmer, E. Wolf, G. Brem, and M. Lipp. 1996. Aputative chemokine receptor, BLR1, directs B cell migration to defined lymphoidorgans and specific anatomic compartments of the spleen. Cell 87:1037.

37. Forster, R., T. Emrich, E. Kremmer, and M. Lipp. 1994. Expression of the G-protein-coupled receptor BLR1 defines mature, recirculating B cells and a subsetof T-helper memory cells. Blood 84:830.

38. Ansel, K. M., L. J. McHeyzer-Williams, V. N. Ngo, M. G. McHeyzer-Williams,and J. G. Cyster. 1999. In vivo-activated CD4 T cells upregulate CXC chemokinereceptor 5 and reprogram their response to lymphoid chemokines. J. Exp. Med.190:1123.

6177The Journal of Immunology

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from

39. Walker, L. S., A. Gulbranson-Judge, S. Flynn, T. Brocker, C. Raykundalia,M. Goodall, R. Forster, M. Lipp, and P. Lane. 1999. Compromised OX40 func-tion in CD28-deficient mice is linked with failure to develop CXC chemokinereceptor 5-positive CD4 cells and germinal centers. J. Exp. Med. 190:1115.

40. Kim, C. H., L. S. Rott, I. Clark-Lewis, D. J. Campbell, L. Wu, and E. C. Butcher.2001. Subspecialization of CXCR5� T cells: B helper activity is focused in agerminal center-localized subset of CXCR5� T cells. J. Exp. Med. 193:1373.

41. Breitfeld, D., L. Ohl, E. Kremmer, J. Ellwart, F. Sallusto, M. Lipp, and R. Forster.2000. Follicular B helper T cells express CXC chemokine receptor 5, localize toB cell follicles, and support immunoglobulin production. J. Exp. Med. 192:1545.

42. Cyster, J. G., K. M. Ansel, K. Reif, E. H. Ekland, P. L. Hyman, H. L. Tang,S. A. Luther, and V. N. Ngo. 2000. Follicular stromal cells and lymphocytehoming to follicles. Immunol. Rev. 176:181–93:181.

43. Gunn, M. D., V. N. Ngo, K. M. Ansel, E. H. Ekland, J. G. Cyster, andL. T. Williams. 1998. A B-cell-homing chemokine made in lymphoid folliclesactivates Burkitt’s lymphoma receptor-1. Nature 391:799.

44. Legler, D. F., M. Loetscher, R. S. Roos, I. Clark-Lewis, M. Baggiolini, andB. Moser. 1998. B cell-attracting chemokine 1, a human CXC chemokine ex-pressed in lymphoid tissues, selectively attracts B lymphocytes via BLR1/CXCR5. J. Exp. Med. 187:655.

45. Tew, J. G., R. M. DiLosa, G. F. Burton, M. H. Kosco, L. I. Kupp, A. Masuda, andA. K. Szakal. 1992. Germinal centers and antibody production in bone marrow.Immunol. Rev. 126:99.

46. Smith, B. A., S. Gartner, Y. Liu, A. S. Perelson, N. I. Stilianakis, B. F. Keele,T. M. Kerkering, A. Ferreira-Gonzalez, A. K. Szakal, J. G. Tew, et al. 2001.Persistence of infectious HIV on follicular dendritic cells. J. Immunol. 166:690.

47. Beadling, C., K. M. Druey, G. Richter, J. H. Kehrl, and K. A. Smith. 1999.Regulators of G protein signaling exhibit distinct patterns of gene expression andtarget G protein specificity in human lymphocytes. J. Immunol. 162:2677.

48. Huang, L., I. Bosch, W. Hofmann, J. Sodroski, and A. B. Pardee. 1998. Tatprotein induces human immunodeficiency virus type 1 (HIV-1) coreceptors andpromotes infection with both macrophage-tropic and T-lymphotropic HIV-1strains. J. Virol. 72:8952.

49. Vissers, J. L., F. C. Hartgers, E. Lindhout, C. G. Figdor, and G. J. Adema. 2001.BLC (CXCL13) is expressed by different dendritic cell subsets in vitro and invivo. Eur. J. Immunol. 31:1544.

50. Shi, G. X., K. Harrison, G. L. Wilson, C. Moratz, and J. H. Kehrl. 2002. RGS13regulates germinal center B lymphocytes responsiveness to CXC chemokine li-gand (CXCL)12 and CXCL13. J. Immunol. 169:2507.

51. Johnson, E. N., T. M. Seasholtz, A. A. Waheed, B. Kreutz, N. Suzuki, T. Kozasa,T. L. Jones, J. H. Brown, and K. M. Druey. 2003. RGS16 inhibits signallingthrough the G�13-Rho axis. Nat. Cell Biol. 5:1095.

52. Sullivan, B. M., K. J. Harrison-Lavoie, V. Marshansky, H. Y. Lin, J. H. Kehrl,D. A. Ausiello, D. Brown, and K. M. Druey. 2000. RGS4 and RGS2 bindcoatomer and inhibit COPI association with Golgi membranes and intracellulartransport. Mol. Biol. Cell 11:3155.

53. Watson, N., M. E. Linder, K. M. Druey, J. H. Kehrl, and K. J. Blumer. 1996. RGSfamily members: GTPase-activating proteins for heterotrimeric G-protein �-sub-units. Nature 383:172.

54. Bouzahzah, F., N. Antoine, L. Simar, and E. Heinen. 1996. Chemotaxis-promot-ing and adhesion properties of human tonsillar follicular dendritic cell clusters.Res. Immunol. 147:165.

55. Moratz, C., V. H. Kang, K. M. Druey, C. S. Shi, A. Scheschonka, P. M. Murphy,T. Kozasa, and J. H. Kehrl. 2000. Regulator of G protein signaling 1 (RGS1)markedly impairs Gi� signaling responses of B lymphocytes. J. Immunol.164:1829.

56. Coico, R. F., B. S. Bhogal, and G. J. Thorbecke. 1983. Relationship of germinalcenters in lymphoid tissue to immunologic memory. VI. Transfer of B cell mem-ory with lymph node cells fractionated according to their receptors for peanutagglutinin. J. Immunol. 131:2254.

57. Olson, T. S., and K. Ley. 2002 Chemokines and chemokine receptors in leuko-cyte trafficking. Am. J. Physiol. 283:R7.

58. Bernstein, L. S., S. Ramineni, C. Hague, W. Cladman, P. Chidiac, A. I. Levey,and J. R. Hepler. 2004. RGS2 binds directly and selectively to the M1 muscarinicacetylcholine receptor third intracellular loop to modulate Gq/11� signaling.J. Biol. Chem. 279:21248.

59. Lippert, E., D. L. Yowe, J. A. Gonzalo, J. P. Justice, J. M. Webster, E. R. Fedyk,M. Hodge, C. Miller, J. C. Gutierrez-Ramos, F. Borrego, et al. 2003. Role ofregulator of G protein signaling 16 in inflammation-induced T lymphocyte mi-gration and activation. J. Immunol. 171:1542.

60. Kosco, M. H., A. K. Szakal, and J. G. Tew. 1988. In vivo obtained antigenpresented by germinal center B cells to T cells in vitro. J. Immunol. 140:354.

61. Kosco, M. H., G. F. Burton, Z. F. Kapasi, A. K. Szakal, and J. G. Tew. 1989.Antibody-forming cell induction during an early phase of germinal centre devel-opment and its delay with ageing. Immunology 68:312.

62. Ansel, K. M., V. N. Ngo, P. L. Hyman, S. A. Luther, R. Forster, J. D. Sedgwick,J. L. Browning, M. Lipp, and J. G. Cyster. 2000. A chemokine-driven positivefeedback loop organizes lymphoid follicles. Nature 406:309.

63. Biberfeld, P., K. J. Chayt, L. M. Marselle, G. Biberfeld, R. C. Gallo, andM. E. Harper. 1986. HTLV-III expression in infected lymph nodes and relevanceto pathogenesis of lymphadenopathy. Am. J. Pathol. 125:436.

64. Fox, C. H., K. Tenner-Racz, P. Racz, A. Firpo, P. A. Rizzo, and A. S. Fauci.1991. Lymphoid germinal centers are reservoirs of human immunodeficiencyvirus type 1 RNA. J. Infect. Dis. 164:1051.

65. Schacker, T., S. Little, E. Connick, K. Gebhard-Mitchell, Z. Q. Zhang, J. Krieger,J. Pryor, D. Havlir, J. K. Wong, D. Richman, et al. 2000. Rapid accumulation ofhuman immunodeficiency virus (HIV) in lymphatic tissue reservoirs during acuteand early HIV infection: implications for timing of antiretroviral therapy. J. In-fect. Dis. 181:354.

66. Hufert, F. T., J. van Lunzen, G. Janossy, S. Bertram, J. Schmitz, O. Haller,P. Racz, and D. von Laer. 1997. Germinal centre CD4� T cells are an importantsite of HIV replication in vivo. AIDS 11:849.

6178 FDC REGULATION OF GC T CELL MIGRATION

by guest on October 16, 2014

http://ww

w.jim

munol.org/

Dow

nloaded from